Benchmark Result Discussion

[Stars&Garters]

What settings did you used with patch? 'Recommended' in the resource page, or mine from here?

Performance Patch (up to 30% more physics speed)

On 5800x3d [threads] computeColliders=6 skinmeshPart=1 affinity=1,3,5,7,9,11,13,15 Before patch After Patch FPS is about the same but physics time is a bit lower.

hub.virtamate.com

I'm running both watercooled CPU and GPU, and i have changed intel's mounting bracket to thermal grizzly one :

Even if my CPU wasn't really thermal throating, hitting at most 80ish degrees in simple physics test before, it's now at max 65 at this test, actually giving a few fps more too.

My RAM is 6400Mhz, but i'm using it at 6000Mhz XMP to reduce VaM crashes. VaM tends to crash very easily with XMP, even if any other games, or RAM benchmarks are just fine.

CPU is at stock settings with just disabled Hyper-threading in bios.

The recommended (all 8s, but with HT enabled and 1,3,5,7,9,11,13,15 affinities), outside of VR at least where I ended up throwing progressively larger numbers (up to 8s) without really knowing what I was doing.

I've tried HT disabled/12345678 affinity a couple of times, without your exact settings admittedly, but it didn't do much. I only have a 3 fan 420mm AIO for CPU cooling and also the Liquid Suprim 4090 but both are obviously not proper custom liquid cooling, and my RAM seems just a tad slower than yours (although I'm also running 6000MHz on an XMP Tweaked profile).

 

[trety]

Seems like you have even beefier cooling than me. I'm not using custom loop, but just 2 AiO systems same like you. And i'm using just 360 for CPU [Kraken z73].
For some reason my system was overperforming other similiar builds somehow, i was at 1.80 physics time in Baseline 3 without the patch, while i saw most of 13900k\4090 systems ~ 2.0.

I would suggest trying my settings without HT. I tried the 'reccomended settings' with HT on, and ended up at 1.56
It was good, sure, but no HT 1-8 are just better -_-
My thinking was, even with the patch, VaM still does love high clocks and it's build on old Unity which wasn't that good with multithread features, so it can really use HT well, like modern games these days. This is one of the reasons why we ended with low physics performance in the first place in VaM 1.x.

 

[Seraphim]

Latest benchmark with cpu patch:

 

[fear_the_potato]

For what it's worth, I decided to do a very simple test. Just loaded up the default scene. Went in to user preferences to see the FPS counter. Moved the camera around violently and watched the FPS. Without patch 12 = ~210fps. With patch 12 = ~310fps.

I loaded up patch 12 on my main VaM installation. It's a very messy, large install. With three animated characters, I went from about 45fps to 60fps. That's a boost!

 

[T1 gumayusi]

Amazing in 1080P!!!

Spoiler: 7500F+4080Super WITHOUT patch 1080P

Spoiler: 7500F+4080Super WITH patch 1080P

Spoiler: 7500F+4080Super WITHOUT patch 2K and WITH patch

desktop 2K,in same time and same scenario,
AVG FPS went from 179 to 262,
GPU utilization from 72% to 98%
GPU Power from 242W to 290W
(4080S Maximum power is 320W)

Spoiler: 7500F+4080Super WITHOUT patch 4KVR and WITH patch

The maximum refresh rate of my headset is 90Hz,so the true change is in baseline3,not the total avg.
from 84.58 to 88.07

 

[DerekH2259]

Nice.
The bloat btw is because EVERY SINGLE addonpackage has it's morphs loaded into the morphbank at start and the morphbank gets iterated through in EVERY SINGLE FRAME. I tried to make all morphs to load ondemand by default, but that broke things.
The morphbank iteration happens when morphs are applied to a character and are counted into "PhysicsTime". So its a burden on the rendering of every character in the scene.

That would explain the massive improvement when running on a clean folder. Is there a known way to mitigate it for those of us with an insane number of .vars, either with changes to the vars if a code fix isn't possible?
I mean I probably ought to clean up, since loading the baseline3 subscene takes a considerable amount of time

 

[fear_the_potato]

That would explain the massive improvement when running on a clean folder. Is there a known way to mitigate it for those of us with an insane number of .vars, either with changes to the vars if a code fix isn't possible?
I mean I probably ought to clean up, since loading the baseline3 subscene takes a considerable amount of time

Maybe this will help? Although I haven't tested it personally.

 

[DerekH2259]

Maybe this will help? Although I haven't tested it personally.

I'd be interested to know how it actually operates. From an initial look, it could be a glamored-up symlink creator

 

[turtlebackgoofy]

I'd be interested to know how it actually operates. From an initial look, it could be a glamored-up symlink creator

yes thats exactly what fixes the bloat

 

[dante101]

i applied this patch and but noting happened....cpu still runs under 20 percent....even in cpu benchmark....i have ryzen1800x.....can anyone help?

 

[Seraphim]

i applied this patch and but noting happened....cpu still runs under 20 percent....even in cpu benchmark....i have ryzen1800x.....can anyone help?

What patch? Should probably ask this in the that patch forum thread.

 

[Mysterio]

Full benchmark, using v12 patch
View attachment 334820

Could you please tell me which settings you use for the CPU Patch?

 

[trety]

Could you please tell me which settings you use for the CPU Patch?

Performance Patch (up to 30% more physics speed)

On 5800x3d [threads] computeColliders=6 skinmeshPart=1 affinity=1,3,5,7,9,11,13,15 Before patch After Patch FPS is about the same but physics time is a bit lower.

hub.virtamate.com

And below are tests i made comapring different settings.
As for Intel, with disabled HT in bios, above gave me best results

Performance Patch (up to 30% more physics speed)

I tried your latest patch 9 but it's crashing on my really old CPU (i7-3770K) during opening the benchmark. Error log: it doesnt support AVX2 so it crashes. You wouldnt get much benefit without AVX2 anyway.

hub.virtamate.com

Performance Patch (up to 30% more physics speed)

I tried your latest patch 9 but it's crashing on my really old CPU (i7-3770K) during opening the benchmark. Error log: it doesnt support AVX2 so it crashes. You wouldnt get much benefit without AVX2 anyway.

hub.virtamate.com

 

[mikesmile]

Here is my benchmark...
which should I upgrade first? GPU or CPU, (memory?)
Name NVIDIA GeForce RTX 2070 SUPER
Processor AMD Ryzen 7 3700X 8-Core Processor, 3600 Mhz, 8 Core(s), 16 Logical Processor(s)
Installed Physical Memory (RAM) 32.0 GB
Ive looked at this thread till my head spins... I hoping someone has seen a similar setup as mine and has had some really significant upgrade experience..
- ive seen a few similar (my speculation on similar intel CPU) where the user upgraded the GPU to 3080 and doubled...

 

[Mysterio]

Performance Patch (up to 30% more physics speed)

On 5800x3d [threads] computeColliders=6 skinmeshPart=1 affinity=1,3,5,7,9,11,13,15 Before patch After Patch FPS is about the same but physics time is a bit lower.

hub.virtamate.com

And below are tests i made comapring different settings.
As for Intel, with disabled HT in bios, above gave me best results

Performance Patch (up to 30% more physics speed)

I tried your latest patch 9 but it's crashing on my really old CPU (i7-3770K) during opening the benchmark. Error log: it doesnt support AVX2 so it crashes. You wouldnt get much benefit without AVX2 anyway.

hub.virtamate.com

Performance Patch (up to 30% more physics speed)

I tried your latest patch 9 but it's crashing on my really old CPU (i7-3770K) during opening the benchmark. Error log: it doesnt support AVX2 so it crashes. You wouldnt get much benefit without AVX2 anyway.

hub.virtamate.com

Thanks! I tried your settings with HT in MSI Click bios 5 turned off but only got slightly better results than HT on and the default settings for my CPU:

[threads]
computeColliders=8
skinmeshPart=1
applyMorphs=8
skinmeshPartMaxPerChar=8
applyMorphMaxPerChar=8
affinity=1,2,3,4,5,6,7,8

[threadsVR]
computeColliders=4
skinmeshPart=1
applyMorphs=2
skinmeshPartMaxPerChar=4
applyMorphMaxPerChar=2
affinity=1,2,3,4,5,6,7,8

[profiler]
enabled=0

Clean install with your settings HT off, I also looked at my taskmanager with HT turned off and it seemed to work correctly,:

Same but with changed skinmeshPart=8

Same but with 3000+ VARs:

Clean install, standard CPU settings, HT on:

Same but with 3000 VARs:

I guess the performance difference comes down to Mainboard and RAM. I don't have anything overclocked.
All benchmarks ran slightly slower with profiler=1.
It's crazy how much the performance tanks in Baseline 3 with the full VAR version =(.

 

[reaper9789]

People keep saying that VaM is cpu dependent, but my cpu usage is usually less than 40% running Vam. The gpu is at 99% at all times.
Maybe the cpu is OP, but still. In VR I struggle to run some scenes.
I didnt get a test done before the mod, But it was worse. I do have 3 tests. Stock clocks, overclocking cpu and overclocking gpu. Overclocking the gpu is by far the biggest difference.
This is 4k resolution on desktop. AMD Ryzen 9 7950x, and AMD RX6900xt with 32gb of ddr5 ram @ 6000mhz.
I am running Windows 11, not sure why it says win 10
Stock clocks

Then CPU overclock Which didnt change much at all.

And now the GPU overclock

Here is a passmark score for reference.

 

[reaper9789]

So... Apparently. one of my CPU power plugs wasnt fully seated. Here is the current result ( GPU still overclocked)

And the passmark difference in CPU performance. its double.

 

[kashh135]

Just got my new desktop with 7800x3d, 4090, 64 GB RAM at 5600, both CPU and GPU water-cooled.

New install:

With performance patch (recommended settings for 7800x3d):

Now, I just need to learn how to create multi-person MMDs to make use of this amazing patch...

 

[driv3rrr]

i am getting low fps ? how do i fix this

 

[starmoon]

i am getting low fps ? how do i fix this

Get better hardware or use desktop mode. VaM needs both, CPU and big GPU performance to do VR and lots of RAM in general if changing assets a lot. A 3060 is will not do well above 1440. The i5 8400 is from 2017 when VaM was first created.

 

[driv3rrr]

Get better hardware or use desktop mode. VaM needs both, CPU and big GPU performance to do VR and lots of RAM in general if changing assets a lot. A 3060 is will not do well above 1440. The i5 8400 is from 2017 when VaM was first created.

damn least 60 i get i will be happy

 

[Vanz]

I'm building a new PC, which graphics card should I get out of these: RX 7900 GRE, RX 7800 XT, 4070, 4070 super?
I'm planning on pairing it with the i5-13600KF. Please enlighten me

 

[Seraphim]

I'm building a new PC, which graphics card should I get out of these: RX 7900 GRE, RX 7800 XT, 4070, 4070 super?
I'm planning on pairing it with the i5-13600KF. Please enlighten me

Get a 7800x3d for $100 more and more than twice the performance in vam, then the card you can afford afterwards. Your performance is guaranteed to be much better than the other pairing.

 

[lolmao500]

Just got my new desktop with 7800x3d, 4090, 64 GB RAM at 5600, both CPU and GPU water-cooled.

How do you buy such a machine and buy DDR5-5600? Like seriously??

 

[Seraphim]

How do you buy such a machine and buy DDR5-5600? Like seriously??

Because its a good stable speed for amd x3D. dont need crazy 8000mhz like intel to run fast.